The aims of this project were 1) to develop a data base for individually housed monkeys consisting of their thrice-yearly psychological assessments along with their health and housing history and 2) to use the data base to test various hypotheses about the occurrence of several kinds of abnormal behavior. The data base consisted initially of psychological assessments (i.e., measures of abnormal behavior and temperament) and various demographic variables such as age, sex, rearing history, room location, and housing situation. Information on housing history (total number of room changes, changes in cage position and cage type) was added to the data base. Medical information was derived from each animal's written health records (e.g., number of lavages, minor surgery, treatments) and also added to the data base. Various hypotheses about the occurrence of self-injurious behavior (SIB) were then evaluated. Monkeys displaying SIB tended to spend a greater proportion of their years in individual cages than normal controls (.91 vs. .80). SIB monkeys were moved more often per year than controls (1.68 vs. 1.57 total moves), but these movements consisted primarily of changes in cage location within the same room. Control monkeys were actually moved to different rooms more often than monkeys with SIB (1.25 vs. .98 room changes). Although a few shipments of monkeys to NERPRC contained a high proportion of monkeys displaying SIB, overall the incidence of SIB was essentially the same in NERPRC born animals and in animals born outside the facility. Differences in early rearing history appeared to have a substantial influence on SIB. Monkeys reared only with other infants in a nursery (peer rearing) were much more likely